CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

external-model-validation

Use when validating an existing prognostic risk signature on an external bulk expression cohort with survival outcomes, producing risk scores, Kaplan-Meier curves, risk distribution plots, heatmap, and time-dependent ROC curves. NOT for: model training, feature selection, nomogram construction, calibration analysis, or single-cell data.

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

81%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable and well-structured: executable commands, concrete input examples, a sequenced validated workflow, and a best-practice navigation table for bundle files. Weaknesses are duplicated scope/ROC-time guidance across sections, CLI detail inlined that belongs in references/cli-guide.md, and broken references to a tests/ directory that is absent from the bundle.

Suggestions

Consolidate scope guidance: merge 'When to Use' / 'When Not to Use' into the Input Validation section (or vice versa) and state the '--roc_times always in years' rule once (the Arguments table) instead of three times.

Move the full Arguments table and usage examples into references/cli-guide.md, keeping only a single quick-start command in SKILL.md, since the navigation table already routes CLI details there.

Fix or remove the dangling tests/ references (tests/data/BRCA_*.csv, tests/testthat.R, tests/refresh_example_output.R, tests/output/) — either bundle those files or point the examples at paths that exist.

DimensionReasoningScore

Conciseness

The body is dense and table-driven with no concept-explaining padding, but scope guidance is repeated across the Input Validation refusal message, 'When to Use', 'When Not to Use', and the Research Use Notice; the '--roc_times always in years' note is stated three times (Arguments, Methods, Examples); and the bundled-test-data command appears twice in Testing. This is more than the 'minor trimmable instances' of level 4, but not the padded/verbose sections of level 2.

3 / 5

Actionability

Fully executable throughout: copy-paste-ready Rscript invocations with all flags, a complete argument table with types/defaults, concrete CSV snippets for all three input formats, error codes with causes and fixes (including a working dedup snippet), and test commands. Common cases (day-unit follow-up, custom colors, ROC positioning) are all covered.

5 / 5

Workflow Clarity

The 4-step workflow is clearly sequenced with explicit validation up front (Step 1 validates files, colors, timeout, seed, time unit), per-step data checks (sample matching, model gene existence), an error-code table that routes persistent failures to references/troubleshooting.md (a feedback loop), and a Testing section with verification commands. Validation is present for this batch operation, so no cap applies.

5 / 5

Progressive Disclosure

The 'When to Read External Files' table provides exemplary, clearly signaled one-level-deep navigation, and every referenced references/*.md and scripts/*.R file exists. Not 5 because the body repeatedly references tests/data/, tests/testthat.R, tests/refresh_example_output.R, and tests/output/ which do not exist in the bundle, and the inlined full CLI argument/usage sections duplicate the dedicated references/cli-guide.md. Not 3 because references are exceptionally well signaled and most content is appropriately placed.

4 / 5

Total

17

/

20

Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is a strong, trigger-rich statement: it names the precise capability, the required input context (external bulk cohort with survival outcomes), the concrete outputs, and a crisp out-of-scope list. The only gap is the absence of synonyms/file-format terms that would push trigger coverage to comprehensive.

DimensionReasoningScore

Specificity

Concrete actions are enumerated: 'validating an existing prognostic risk signature' and 'producing risk scores, Kaplan-Meier curves, risk distribution plots, heatmap, and time-dependent ROC curves'. Coverage is comprehensive for the domain, exceeding the minor-gaps level of 4, and the voice is appropriate (no first/second person).

5 / 5

Completeness

Explicitly answers both what ('validating an existing prognostic risk signature... producing risk scores, Kaplan-Meier curves...') and when ('Use when validating... on an external bulk expression cohort with survival outcomes') with concrete triggers, plus an explicit NOT-for boundary — matching the anchor-5 example pattern.

5 / 5

Trigger Term Quality

Natural domain phrases users would say are present: 'prognostic risk signature', 'external... cohort', 'survival outcomes', 'Kaplan-Meier', 'ROC curves', 'bulk expression'. Falls short of 5 because no file extensions (e.g., CSV) or common synonyms (e.g., 'gene signature', 'survival analysis') are included.

4 / 5

Distinctiveness Conflict Risk

A clear niche (external validation of a fixed prognostic model) is reinforced by an explicit exclusion list — 'NOT for: model training, feature selection, nomogram construction, calibration analysis, or single-cell data' — giving minimal conflict risk with adjacent biostatistics skills.

5 / 5

Total

19

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

aipoch/medical-research-skills

Path

awesome-med-research-skills/Data Analysis/external-model-validation/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.